AN EXPERIMENTAL INVESTIGATION OF FIRE PERFORMANCE OF EARTHQUAKE DAMAGED STRUCTURES

Abstract

Fire is possibly the most severe environmental hazard to which the built infrastructure might be subjected to. The risk of fires in the aftermath of earthquakes is a credible extreme load in seismic regions around the world and such events are considered a major threat to life and property. A very important aspect of the research on the concrete structures exposed to elevated temperatures is the study of impact of fires on the structures pre-damaged by earthquakes. Over the past centuries fire has emerged as an integral part of emergency response strategies which are focused on life safety as well as the infrastructure safety of any nation. However no current regulations consider the fire and earthquake hazard in a sequential manner. The overall behaviour of reinforced concrete structures, during the time it is exposed to fire and even after, is still a hot issue in civil engineering. As far as the behaviour during fire is concerned, the human safety and the structural resistance of the structures are two important associated interests involved while as the behaviour of the structure after fire is related to the residual bearing capacity of the structure. In spite of the fact that the structural materials undergo mechanical decay after a thermal accident, it is generally necessary to ascertain and quantify the residual capacity of a structure and compare it with the safety levels. A critical literature review reveals that a number of studies have been carried out on the different aspects of concrete structures at elevated temperatures. However no significant research attempt has been made in the past for assessment of structures under fire following earthquake. A critical literature review mandates consideration of such loading conditions under performance based design techniques. Since fire following earthquake falls out of scope of practising professionals in earthquake and fire service fields, it has stimulated the interest amongst various groups of researchers in the structural engineering community to carry out a rigorous investigation in this regard. Another important observation from the literature review suggests that the past studies have been mostly carried on the ductile detailed concrete structures. This has led to a information gap with regards to the behaviour of non-ductile concrete structures, built before 1980’s, which are still more in number than the ductile ones. Most of the design codes specify the fire rating, i.e. the time to failure of a structural member exposed to fire, as the measure of the building elements to resist fire. However the biggest drawback of these codes is the approach of relating the size and the cover to reinforcement as the only parameters to determine the fire endurance of the structural elements. The effect of the iii concrete strength, the tie configuration and the percentage confining reinforcement is completely neglected in determining the fire ratings in reinforced concrete columns. The effect of confining reinforcement on the fire resistance of RC columns is not well documented. No design tools are available in the literature which takes into consideration the effect of confinement on the fire ratings of RC columns Thus the research efforts needed to fill the gaps in the state of art may be put at three different levels: structural, elemental and material. Structural level represents the holistic behaviour of a framed structure consisting of beams, columns, slabs and joints. Elemental level represents the behaviour of full scale individual structural elements like columns, beams etc whereas, material level constitutes the studies on the behaviour of constituent materials such as steel and concrete. The principal objective of the current research was to experimentally obtain the behaviour of a full scale RC frame with non ductile detailing with and without infill. At elemental level, a study was carried out to study the effect of degree of confinement and concrete strength on fire resistance of reinforced concrete columns. In order to evaluate the behaviour of a reinforced concrete (RC) frame in post earthquake fire, a full-scale RC frame assemblage was constructed without following the ductile detailing guidelines of the Indian standards. A three phase test procedure was adopted in testing the RC frame. The frame was first subjected to a cyclic lateral load to simulate the seismic effects which was then followed by a one hour compartment fire. After fire the frame was subjected to a residual cyclic load test. The cyclic lateral load was applied on the RC frame using two double acting hydraulic actuators acting in tandem with each other against a strong reaction wall. The main aim of the mechanical loading was to induce a damage corresponding to the collapse prevention performance level (S5) as prescribed by FEMA 356:2000. Following the lateral seismic load test, the RC frame was subjected to a full scale fire test wherein a designed compartment fire was developed using kerosene as the fire source. After exposing the RC frame to compartment fire, the RC frame was again tested under lateral seismic load for measuring the residual lateral load capacity of the frame. Numerous sensors, namely thermocouples, strain gauges and linear variable differential transducers were embedded at key locations of the frame to capture important information during the tests. The results show a marked influence of reinforcement detailing on the post-earthquake fire performance of the concrete structures. The simulated earthquake loading caused wider cracks and more severe concrete spalling in the frame without ductile detailing compared to the frame with ductile detailing. Spalling of iv concrete and buckling of reinforcement was observed in top beams. The present study reveals the conservativeness of the damage levels specified in Table C1-2 of FEMA 356 (2000) vis-àvis both the RC frames (ductile and non-ductile). The overall damage in the test frame was not severe as anticipated in the Table C1-2. However, the damage caused in the non-ductile detailed frame was in line with the expected damage specified by C1-3 of FEMA 356. Present study also reveals overestimation of permanent drifts given in FEMA 356. However, the permanent drift in non-ductile detailed RC frames is higher than in ductile detailed frames. The investigation validates the time-temperature curve designed according to the fire design equation of Thomas and Heseldon (1972). The tests reveal the vulnerability of thin elements like slabs and shells of non-ductile RC frames to spalling in post-earthquake fire events. While drawing the attention towards addressing the issues of fire following earthquakes, the melting of reinforcement after spalling of thin elements like slab and shells needs to be considered in design codes. The test reflects the better performance of the RC frames with ductile detailing. Thus, the recommendations generally used in the seismic resistant design may also be helpful in enhancing the fire resistance of the RCC structures. The next component of the dissertation was carried out to study the effect of the masonry infill on the response of an in-filled RC framed building in post-earthquake fire. Same testing procedure, as in earlier component, was followed in this study also. The results reveal a better performance of the frame in the cyclic load test with higher load carrying capacity at S5 level than the bare frame. Cracks were developed on the in-plane infill walls, though the masonry walls remained intact after the cyclic load test. The plinth beams, along the loading direction, however got damaged with a hinge formation and buckling of reinforcement. This study indicates that for a frame that is properly designed for seismic loads, infill panels will most likely have a beneficial influence on its performance. The study reveals that infill panels can be used to improve the performance of existing non-ductile frames. The masonry infill helps in delaying the flashover. The time taken to reach the maximum temperature is higher in masonry in-filled frames. The brick infill walls provided insulation to the RC structural elements and slowed the transmission of heat to these elements. This beneficial effect of masonry walls should be considered while designing the columns and beams, which are integrated in the masonry walls. Finally, towards the end, the effect of the degree of confinement, tie configuration and concrete strength on the fire resistance of the RC columns was studied. A total of eight full scale columns of dimensions 2800×300×300 mm were cast with different confinement levels v achieved by changing the spacing of the traverse steel reinforcement. A full scale column fire furnace was designed and built to test the columns under a standard fire curve of ISO 834:1975. A standard testing procedure was followed where in the columns were axially loaded to the 40% of the ultimate load carrying capacity, 60 minutes before exposing them to the fire. The fire was continued till the columns failed to take load. The results show that the degree of the confinement has a marked role to play in the fire rating of the reinforced concrete columns. With the increase in the confining reinforcement factor, the fire resistance of the RC columns gets enhanced. The effect of the confining reinforcement is more pronounced in NSC columns than in HSC columns. Higher confinement also prevents the spalling of the higher strength concrete. A design equation has also been developed to measure the fire resistance of RC columns which also takes confinement parameters into account. The design equation can be integrated in the design standards to facilitate the fire design of columns.